gradle-schemaspy-plugin

A plugin for Gradle to allow for running SchemaSpy as a part of the buildDashboard plugin.

Purpose

This Plugin allows for running SchemaSpy against a database as part of a Gradle build. SchemaSpy's output will be accessible via Gradle's BuildDashboard Plugin.

Instructions

Add a file named schemaspy.properties in the config\schemaspy directory under your Gradle project's root. That file should contain key=value pairs for all required SchemaSpy arguments as detailed in the Running SchemaSpy section of the SchemaSpy documentation.

For example, for a Postgre database, your schemaspy.properties file should look something like this:

t=pgsql
db=TomsDB
s=public
u=UserForTomsDB
p=SecretPassword
dp=C:/Users/Tom/.gradle/caches/modules-2/files-2.1/postgresql/postgresql/9.1-901-1.jdbc4/9bfabe48876ec38f6cbaa6931bad05c64a9ea942/postgresql-9.1-901-1.jdbc4.jar
host=localhost
port=5432 
o=C:/Users/Tom/Programming/Projects/TestGradleSchemaspyPlugin/build/reports/db/schemaspy

You can use Gradle to substitute into the file the path to the downloaded jar of the dependency pretty easily using the configurations.derby.asPath value with Gradle's filter method:

filter(org.apache.tools.ant.filters.ReplaceTokens, tokens: ['driver_path_token':'path_to_driver'])

In your build.gradle file, pass the schemaspy configuration dependency on the database driver for the database on which you wish to report.

For example, for a Postgre database, your build.gradle should look something like this:

configurations {
  postgres
}

dependencies {
  postgres(group: 'postgresql', name: 'postgresql', version: '9.1-901-1.jdbc4')
  
  configurations.schemaspy.dependencies.addAll(configurations.postgres.dependencies)
}

And then to link to SchemaSpy's output (and any other Reporting tasks you're interested in) via the generated Build Dashboard:

gradle.taskGraph.whenReady { taskGraph ->
  // Check to add Reporting tasks to the Build Dashboard
  if (taskGraph.allTasks.any { Task t -> t instanceof GenerateBuildDashboard }) {
    List<Reporting> reportsToAggregrate = []
	
    // Add Schema Spy to build Dashboard (if it has been run)
    def schemaSpyTasks = taskGraph.allTasks.findAll { Task t -> t instanceof SchemaSpyReportTask }
    reportsToAggregrate.addAll(schemaSpyTasks)
    
    // ... aggregate any other Reporting tasks to the Dashboard here
    
    // Aggregate all aggregatable reports to be run to Dashboard
    if (reportsToAggregrate) {
      buildDashboard {
        aggregate(reportsToAggregrate as Reporting[])
      }
    }
  }
}

// Define ordering of tasks to ensure reports are generated prior to being accumulated by build dashboard
buildDashboard.mustRunAfter schemaSpy

Extras

Included in the project is a net.sourceforge.schemaspy.SchemaAnalyzer implementation for Apache Derby running in file system mode (no support for in-memory databases currently).

One cavaet for running this plugin against a Derby DB on Windows is the need to quadrupal-escape any path separator characters in the Schemaspy db argument within the schemaspy.properties file.

dp=C:\\\\Users\\\\Tom\\\\Programming\\\\Projects\\\\TestDerbyProject\\\\testderbydb